Schizophrenia is generally regarded as a diverse and heterogeneous disorder and it has been difficult to establish meaningful subtypes that can be reliably identified, that have a stable course, and have demonstrable neurobiological underpinnings. In previous work, we and others have identified a subgroup of schizophrenic patients with severe, generalized cognitive impairment, a very poor outcome, predominance of negative, or deficit, symptoms, and a history of unremitting, chronic illness. Patients in this subgroup have been characterized as "Very Poor Outcome" or "Kraepelinian" schizophrenics. Another subgroup of patients has been identified that meet all diagnostic criteria for schizophrenia but have normal or only mildly impaired cognitive function. These patients have a more benign disorder characterized by a relatively good outcome, and a history of intermittent episodes of relapse involving re-emergence of positive symptoms, intermixed with varying lengths of stable life in the community. We and others also have identified distinct molecular abnormalities related to membrane phospholipid and high-energy phosphate metabolism in never-medicated, first-episode schizophrenics thought to reflect exaggerated synaptic pruning. Our recent findings indicate these molecular changes are more prominent in subjects with cognitive impairment. The proposed research will evaluate the general hypothesis that schizophrenia may be divided into two major syndromes differing in clinical status, course, neurobiological and molecular correlates. We propose to utilize clinical and neurocognitive assessment in association with 31P-1H magnetic resonance spectroscopy imaging (31P-1H MRSI) and magnetic resonance imaging (MRI) to evaluate this general theory. Our hypothesis is that schizophrenics with cognitive deficit will have MRSI and MRI findings consistent with a marked exaggeration of synaptic pruning and schizophrenics with no or mild cognitive deficit will have no or minimal MRSI findings.